Wikipedia:Articles for deletion/Topic-based vector space model
This looks to me as original research. The only reference in this article is the paper Topic-based vector space model. This article basically summarizes that paper, which is published in 2003. As such, I would say it is unecyclopedic. Oleg Alexandrov 01:00, 10 Jun 2005 (UTC)
- Delete at best, this is highly specialized topic that is unlikely to amount to much more than gibberish to the non-engineering/mathematician reader. In effect, it's a technical DicDef that fails in clarity. In my considered opinion as an Engineer, the cited VSM article, tho' better written, should probably accompany it, for same reasoning. Fabartus 03:43, 10 Jun 2005 (UTC)
- Delete, but for a different reason. I don't mind specialized articles; its just that this one is so incredibly poorly written. And mind you, it was written by the very same person (Dominik Kuropka) who wrote the journal article in 2003. If the original author can't be bothered to state their claim clearly, why should we bother? linas 04:27, 10 Jun 2005 (UTC)
- Expand, a weak keep I have tried to clean the article up a bit, so that it is at least parsable. I certainly wouldn't shed any tears over it if it fell off a cliff, but at least it might be worth keeping long enough to see if someone can expand it. -Phantym 04:54, 10 Jun 2005 (UTC)
- Merge (and redirect) as a little note into the VSM article, and expand that. Pcb21| Pete 07:50, 10 Jun 2005 (UTC)
- Merge into vector space model. As far as I understand it, the vector space model is one of the traditional models in information retrieval, and definitely deserves an article in Wikipedia. The topic-based VSM seems barely notable, but it is published in conference proceedings, so we can mention it in the VSM article. I doubt that lack of clarity is a reason for deletion. Jitse Niesen 09:34, 10 Jun 2005 (UTC)
- Merge with redirect to vector space model. I haven't messed with serious math at greater than high school level for over twenty years, but this article and the accompanying references are quite clear and informative about their subject matter. You transform a document into a series of multidimensional arrays ("vectors in a multidimensional space") which can then be manipulated algorithmically by just about any automatic filtering method ever invented. The reference paper provides a codable schema and some simple examples in SQL to enable similarity metrics to be computed on documents. --Tony Sidaway|Talk 13:22, 10 Jun 2005 (UTC)